Listen to Semantic Scholar Papers
CastReader reads Semantic Scholar abstracts, TLDRs, and paper intros aloud — perfect for scanning a week's worth of new papers on the commute.
Scaling test-time compute improves reasoning in small language models
Five years ago, Port Newport handled fewer than 40 cargo vessels a week. Today it handles more than 180, and almost none of that growth came from new docks or cranes. The change was almost entirely in the data layer — a shift from faxed manifests to a shared real-time scheduling system. What used to take four hours of phone calls now takes four minutes in a web dashboard.
Reads the abstract and intro — citation counts and paper chrome are skipped
How CastReader Works with Semantic Scholar
Built for paper-triage workflows
TLDR First
🔊 Semantic Scholar's AI Summary Read First
Semantic Scholar auto-generates a TLDR for most papers. CastReader reads the TLDR first so you can triage in 30 seconds — then continues to the full abstract and intro if you want more depth.
Citation Graph Skipped
No 'Cited by 284 · References 67'
Semantic Scholar pages are dense with citation graphs, 'related papers', and 'influential citations' widgets. CastReader skips all of that chrome — you hear the paper's own words, not the metadata.
Author List Handled
'Chen, Shazeer, Zoph et al.' — Not Read
Long author lists (15+ co-authors on ML papers is normal) get skipped by default. CastReader reads the paper title and goes straight to content. Authors stay visible for when you want them.
Equation Announcements
LaTeX / MathJax Announced, Not Read
Inline equations get announced ('equation') and skipped. Reading LaTeX aloud is nonsense. The text around the equation flows naturally.
Frequently Asked Questions
Everything you need to know about listening to Semantic Scholar with CastReader
Does Semantic Scholar have a built-in read-aloud feature?
No. CastReader adds one for free on every paper page.
Does it read the full paper or just the abstract?
By default: TLDR first, then abstract, then intro. Full-body reading is available if you click Listen deeper in the paper — CastReader handles section headers cleanly.
Does it read citations and references?
No. The references list and inline citation markers ([12], [Smith 2023]) are skipped. The surrounding prose reads naturally.
Does it work on PDF views?
Yes. Semantic Scholar's embedded PDF viewer is supported — CastReader reads the rendered text. For heavy PDFs with complex layouts, use /pdf-to-audiobook for better extraction.
Is it free?
Completely free. No API key, no usage limits, no account required.
Also Try
Listen to arXiv
arXiv preprints read aloud
Listen to ResearchGate
ResearchGate papers read aloud
Listen to SSRN
SSRN working papers read aloud
Listen to PubMed
PubMed biomedical abstracts read aloud
Listen to JSTOR
JSTOR humanities and social science papers
PDF to Audiobook
Upload full PDFs for complete paper audio
Start Listening Now
Completely free. No signup. Install and start listening to Semantic Scholar papers.